University of Amsterdam
Amsterdam Business School
MSc Finance, Corporate Finance
Master Thesis
Does overconfidence impact the returns of serial acquirers?
by
Christakis Kourounis
July 2017
2
Statement of Originality
This document is written by student Christakis Kourounis who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 Table of Contents 1. INTRODUCTION ... 4 2. LITERATURE REVIEW ... 5 3. METHODOLOGY ... 9 3.1 Event study ... 9
3.2 Hypothesis and models ... 10
4. DATA AND DESCRIPTIVE STATISTICS ... 12
4.1 The sample ... 12
4.2 Descriptive statistics ... 13
5. RESULTS ... 18
5.1 CAR’s of serial acquirers ... 18
5.2 Regression analysis ... 20
6. CONCLUSION ... 25
4
1. Introduction
The performance of acquiring firms and the gains of acquisitions to shareholders, have been a popular field of research in the past decades. Findings of prior literature agree that, while acquiring public targets is associated with wealth reduction for the shareholders of bidding firms, as reported by Fuller et al. (2002), the acquisition of private targets is associated with better results and wealth creation, Moeller et al. (2005) and Andrade et al. (2001). Studies of Conn et al. (2005) and Croci & Petmezas, (2009) on serial acquirers, have shown that, for such acquirers, the abnormal returns show a declining trend from one deal to the next. The reasons might be due to declining investment opportunities on later deals or due to the market anticipation effect for acquisition programs as described by Klasa & Stegemoller, (2007), and Schipper and Thompson (1983) respectively.
This thesis adds to the literature, by examining the effects of overconfidence on acquisition announcement returns for serial acquirers, and the decision between methods of payment. There is a key distinction between overconfidence and empire building behavior: while empire builders are more concerned about personal benefits and control, overconfident managers try to act in the best interest of their shareholders regardless of their beliefs being irrational. This is indicated by their over-investment in their own firms, by not exercising in-the-money vested stock options, as Malmendier and Tate (2008) argue. CEO’s might become overconfident following the implementation of a successful deal, a phenomenon known as self-attribution bias, or might be overconfident by nature, leading them to poor choices in future investments in either case, as described by Billet & Qian, (2008). It also contributes to the existing literature by examining the notion of managers learning from experience, described by Aktas et al. 2010: The markets reactions triggered from an investment project contain information that the management should use to “learn from experience” and not repeat the
5
same mistakes. Following a bad deal, the CEO should re-evaluate his beliefs and adjust future bids accordingly.
First, I find that findings are consistent with previous evidence of negative abnormal returns for bidders that acquire public targets, and of the declining trend these abnormal returns for serial acquirers. Second, I find a positive correlation between the returns of consecutive deals. However, I do not find any evidence that CEO’s overconfidence either strengthens or weakens this correlation. Finally, I find that the likelihood of choosing cash as the method of payment is not affected by overconfidence. Overconfident managers don’t seem to prefer to pay by cash, as suggested by the idea that external funds are perceived as costly due to under-valuation of their firms. Overall the results can be interpreted as a contradiction to the learning hypothesis: neither rational, nor overconfident CEO’s seem to evaluate correctly the investors’ signals, reflected on the announcement of previous bids, to “learn by experience” and achieve better results in their subsequent acquisitions.
2. Literature review
While consensus has been reached for the benefits of mergers and acquisitions to the shareholders of target firms, despite the vast amount of literature and the empirical evidence regarding the returns to the shareholders of acquiring firms is not conclusive: some researchers report a negative effect on stock price in the short-run, while others report a positive effect. Specifically, on one hand, several studies have shown that acquiring firms experience, on average, statistically significant negative abnormal returns (Fuller et al. 2002; Betton, Eckbo & Thorburn, 2008; Hackbarth & Morellec, 2008; Harford, Jenter & Li, 2011; Doukas et al. 2002; Goergen & Renneboog, 2004; Morck, Schleifer & Vishny, 1990; Croci & Petmezas, 2009; Chang, 1998; and Doukas and Petmezas, 2007). The reasons as to why significant losses
6
occur for acquiring shareholders has been given many interpretations: One possible explanation comes from the free cash-flow hypothesis, described by Jensen (1986) and Harford, (1999). It suggests that empire building managers in firms with large cash reserves are more prone to undertake value destroying investments (if no value enhancing investments are available), than to give dividend payouts to shareholders. In the same notion, CEO’s may show hubristic behavior that leads to value destructive acquisitions by overpaying for the target as reported by Roll (1986). An alternative idea is presented by Malmendier and Tate (2008) who argue that overconfident managers, unlike empire builders, actually believe that are acting in the best interest of the shareholders when undertaking acquisitions. They overestimate their ability to create value, both internally and externally, and believe that the market misvalues their projects, leading them to undertake acquisitions with lower or even negative returns. Further, negative abnormal returns for firms that finance acquisitions by stock, could be explained by the equity signaling hypothesis. The market might perceive the stock financing decision as a signal that the company is overvalued, as Travlos (1987) and Myers & Majluf (1984) suggest. Intuitively follows that investors would prefer to pay low rather than high for an overvalued stock, leading to negative announcement abnormal returns for the bidder.
On the other hand, some studies have found positive and significant abnormal returns for the shareholders of acquiring firms, both on domestic markets (Doukas et al. 2002; Fee & Thomas, 2004; Kohers & Kohers, 2000; Maquieira, Meginson & Nail, 1998; Schwert, 1996, Cakici et al. 1996) as well as in cross-border M&A deals Martynova & Renneboog, (2008). The positive short-term returns are mostly clustered around private targets as shown by Netter & Stegemoller, 2002; Moeller, Schlingemann & Stulz, 2005; Andrade et al. 2001 and Masulis, Wang & Xie, 2007). One possible explanation for the positive returns of private firms is that they are less liquid than the public ones, and according to Haleblian et al. (2006) and Officer et al. (2008) this illiquidity will be reflected in their valuation after the acquisition process. An
7
alternative explanation offered is that, if the acquisition of private firms is financed by equity, it leads to the creation of outside block-holders. These block-holders can play a crucial role by closely monitoring of the management performance, which leads to higher bidder returns for these kind of deals as suggested by Shleifer & Vishny, (1986) and Chang (1998). Another possible explanation comes if we take into consideration the tax benefits of stock acquisitions. According to Ismail (2008), owners of privately held companies enjoy a tax deferral option when they are paid with equity, and if this option is in-the- money they are more willing to accept a lower price to complete the deal. This in turn will lead to higher abnormal returns for the acquirer.
A large body of literature has also addressed the issue of short-term returns with regard to “serial acquirers”: firms that tend to frequently acquire other firms. The findings about the evolution of announcement returns for these firms are again puzzling: Some authors identify returns to be lower or negative in subsequent deals, compared to the first deals, and others present evidence of positive announcement returns in later deals. More specifically, on one hand, studies on the U.S. market from (Fuller et al. 2002; Conn et al. 2005; Croci & Petmezas, 2009; Ahern, 2008; Ismail, 2008, Billet & Qian, 2008, Boubakri, Chan & Kooli, 2012) show that the cumulative abnormal returns of serial acquirers tend to have a declining trend from deal to deal. The learning hypothesis suggests that management should take into consideration the information from the market’s reaction incorporated into the stock price during the previous deals, and based on this information adjusts accordingly its future investment decisions. After a successful deal, due to self-attribution bias, the management may become overconfident, or infected by hubris. In this context, more experience and better past acquisition performance can lead the CEO to take irrational decisions and engage in value destructive subsequent acquisitions according to Conn et al. (2005), Ismail (2008), and Billet & Qian, (2008). In contrast, Aktas et al. (2010) suggest that managers are either rational or overconfident/hubristic
8
by nature, and that the latter actually benefit from learning and improve their results from deal to deal. Another possible explanation for the declining abnormal returns is given by Ahern (2008) who argues that after multiple acquisitions, firms not only are bigger themselves but also go after bigger targets. During an acquisition, this leads to very increased costs that “cut” a bigger portion of the perceived benefits. Among other possible explanations are the Diminishing Returns Hypothesis and the declining investment opportunity of Klasa & Stegemoller, (2007) which suggest that returns of later acquisitions are expected to be worse than the returns of previous acquisitions as the best investment opportunities are already implemented. Schipper and Thompson (1983) pointed out that acquisitions are a repetitive process. Based on their acquisition program anticipation theory and the Capitalization Hypothesis it is expected that later deals will have insignificant returns since the value of subsequent acquisitions are already capitalized by the investors at the announcement of the acquisition program, while Fuller et al. (2002) argue that declining returns may be a result of ownership dilution effect for public targets and for stock financed deals.
Contrary to the above, studies by Asquith et al. (1983) and Croci & Petmezas (2009), showed that serial acquirers experienced positive and significant announcement abnormal returns through the higher order deals. These results are in line with the Organizational Learning theory which suggests that taking into consideration the learning curve and accumulation of experience through the process of multiple acquisitions, the abnormal returns of subsequent acquisitions should be increasing. Active learning by frequent acquirers can provide a competitive advantage, enabling the management to develop expertise in identifying good targets and being able to integrate them through successful restructuring, creating value for the shareholders, as argued by Croci & Petmezas (2009), who highlight the importance of the managerial quality in their work.
9
3. Methodology
Similar to Malmendier and Tate (2008) a CEO in my sample is classified as overconfident based on his over-investment in the firm. More specifically, the executive is labeled as overconfident if he does not exercise stock options that are vested with value at least 50% higher than the value of the option at grant date. This is the key difference between overconfidence and empire-building behavior that leads to value destructive investments. An overconfident manager believes he is doing the best for the shareholders that is why he over-invests in his firm. This does not seem to be the case for empire builders who can be argued that are more concerned with maintaining their personal benefits of control rather than taking decisions that aim at increasing shareholders wealth.
3.1 Event Study
Using the Event Study methodology, I calculate the abnormal returns, for the acquiring firm, centered at the date of the acquisition announcement as this is reported by Thomson One, as follows:
𝐴𝑅𝑖,𝑡 = 𝑅𝑖,𝑡− (𝛼𝑖 + 𝛽𝑖𝑅𝑚,𝑡)
where 𝐴𝑅𝑖,𝑡 are the abnormal returns of stock “i” in time “t”, and are calculated as the actual observed returns 𝑅𝑖,𝑡, less the estimated normal returns predicted by the market model (𝛼𝑖 + 𝛽𝑖𝑅𝑚,𝑡). Following Moeller et al. (2005), the estimation period for the expected returns is set to 200 trading days, ending 6 days prior to the deal announcement.
Next, I aggregate the daily abnormal returns over the three-day event window to arrive at the cumulative abnormal returns:
𝐶𝐴𝑅(−1, +1) = ∑ 𝐴𝑅𝑖,𝑡 +1
10
3.2 Hypothesis and models
Hypothesis 1: The abnormal returns of current acquisition announcement are positively correlated to the announcement returns of the previous acquisition announcement and this correlation is stronger with overconfident managers.
I estimate an OLS regression model to test my first hypothesis:
𝐶𝐴𝑅𝑖,𝑇 = 𝛽𝜊+ 𝛽1 𝐶𝐴𝑅𝑖,𝑇−1 + 𝛽2 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖 + 𝛽3𝐶𝐴𝑅𝑖,𝑇−1∗ 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖+
𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠 (1)
where 𝐶𝐴𝑅𝑖,𝑇 is the cumulative abnormal return for stock of firm “i” at the announcement of the most recent (current) acquisition and 𝐶𝐴𝑅𝑖,𝑇−1 is the cumulative abnormal return for stock of firm “i” at the announcement of its previous acquisition. 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖 is an indicator variable equal to one, if the CEO is classified as overconfident. The coefficient 𝛽1 in specification (1) can be interpreted as the correlation coefficient of subsequent bid announcement returns to past bid announcement returns. I expect this coefficient to have a positive value to support the hypothesis that abnormal returns of serial acquirers exhibit persistence from deal to deal. Coefficient 𝛽2 captures the effect that overconfident CEO’s have on the announcement returns. A negative sign would suggest that overconfidence leads to deteriorating returns and wealth destructions rather than creation. The interaction term (𝐶𝐴𝑅𝑖,𝑇−1∗ 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖) is included to assess the joint effect of overconfidence and past acquisition performance, whether overconfident managers increase the effect past returns have on future returns.
Hypothesis 2: With an overconfident CEO, a firm is more likely to continue having negative announcement abnormal returns in the subsequent acquisitions if it experienced negative abnormal returns in the previous acquisition.
11
To capture the effects of bad past performance on the likelihood of realizing bad returns on future bids I estimate the following Logit regression model:
Logit{𝑁𝑒𝑔 (𝐶𝐴𝑅𝑖,𝑇) = 1} = 𝛽𝜊+ 𝛽1𝑁𝑒𝑔 (𝐶𝐴𝑅𝑖,𝑇−1) + 𝛽2 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖+
𝛽3𝐶𝐴𝑅𝑖,𝑇−1∗ 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖+ 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠 (2)
where 𝑁𝑒𝑔 (𝐶𝐴𝑅𝑖,𝑇) is an indicator variable equal to one if the firm experienced negative abnormal returns when it announced its current acquisition and 𝑁𝑒𝑔 (𝐶𝐴𝑅𝑖,𝑇−1) is an indicator equal to one if the firm had negative abnormal returns at the announcement of its previous acquisition. 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖 is an indicator variable equal to one if the CEO is classified as overconfident. Positive (negative) coefficients on specification (2) would suggest firms run by overconfident management and experienced negative returns of past announcement are more (less) likely to experience again negative returns in the next deal announcement.
Hypothesis 3: Overconfident CEO’s are more likely to finance acquisitions by cash, especially if they manage cash rich firms.
To capture the effect of overconfident managers on the choice of payment for a deal I estimate the following Logit regression model.
Logit{𝐶𝐴𝑆𝐻𝑖,𝑡 = 1} = 𝛽𝜊+ 𝛽1 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖 + 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠 (3)
where, 𝐶𝐴𝑆𝐻𝑖,𝑡 is a dummy equal to one if the firm paid for the acquisition using only cash and 𝑂𝑣𝑒𝑟𝑐𝑜𝑛𝑓𝑖𝑑𝑒𝑛𝑡𝑖 is an indicator variable equal to one if the CEO is classified as overconfident. If has been argued that overconfident managers are more likely to use cash as medium of payment for the acquisitions they undertake, especially in firms with abundant internal resources. I would expect the coefficient 𝛽1 to have a positive sign.
Across all specifications, I use several controls variables reported by related literature to have an impact on the short-term abnormal returns of an acquisition announcement. Among them,
12
high Tobin’s Q, equity financing, firm size, cash abundance, low leverage and cross-industry deals are factors associated with lower returns.
4. Data and Descriptive Statistics 4.1 The sample
The data sample starts on January 1, 1992, since Compustat-Execucomp does not have information about executive’s compensation prior to this date, and ends December 31, 2016. Information about the deals is extracted from Thomson One database and includes publicly traded firms that during the sample period have successfully acquired at least two publicly traded targets. As in Fuller et al. (2002) and Cai & Vijh, (2007), financial firms are not included in the sample due to special characteristics such as the regulatory environment and high leverage levels. Further, by applying similar restriction as in Moeller, Schlingemann, & Stulz, (2005) and Phalippou et al. (2015), the final sample contains only successful acquisitions with minimum deal value of $1 million, the acquirer holds 100% of the target’s shares after the acquisition, while prior to the deal it held less than 50% and both acquirers and targets are U.S. based firms. Data on stock is gathered from the Center for Research in Security Prices (CRSP) and is further complemented with several accounting items from the CRSP-Compustat-Merged database. Finally, from Compustat-Execucomp I gathered data regarding compensation and stock option grants for the Chief Executive Officers, which are used to construct the Overconfident-CEO’s proxy. Following a somewhat similar approach to Malmendier & Tate (2008), I characterize a CEO overconfident if during his tenure he holds exercisable stock options with value at least 50% higher than the value of these options at grant date.
13
4.2 Descriptive Statistics
Table 1 presents summary statistics for deals and for executives for the whole sample, as well as by deal order. The most recent acquisition made by the acquirers in the sample period is expressed as 3rd deal, while the first acquisition (the most distant in time) is expressed as 1st deal. Diversified indicates whether the target belongs to the same industry group as the acquirer based on 2-digit Standard Industrial Classification (SIC) codes and the variable Overconfident is used to classify a CEO as overconfident if he holds exercisable stock options even though the option’s value is at least 50% higher than its value at grant date.
Table 1.
The table shows summary statistics of deals (Panel A) and executives (Panel B) for the whole sample, and by deal order. 1st deal denotes the first deal done by an acquirer in the sample period and 3rd deal denotes the last (most recent) deal. Only cash, Only stock and Hybrid & Other indicate whether the financing method was pure cash, pure stock or mixed respectively. Diversified (Non-diversified) indicates whether the target belongs to a different (same) industry group as the acquirer based on 2-digit Standard Industrial Classification (SIC) codes. Overconfident is an indicator equal to one if a CEO holds exercisable stock options even though the option’s value is at least 50% higher than its value at grant date.
Panel A: Deal Characteristics
Whole sample 1st Deal 2nd Deal 3rd Deal No. Frequency No. Frequency No. Frequency No. Frequency
Only Cash 576 42.8% 173 36.7% 192 40.7% 87 45.5%
Only Stock 300 22.3% 123 26.1% 105 22.2% 42 22.0%
Hybrid & Other 470 34.9% 176 37.3% 175 37.1% 62 32.5% Non-diversified 871 64.4% 305 64.6% 287 60.8% 125 65.5%
Diversified 482 35.6% 167 35.4% 185 39.2% 66 34.6%
Deals 1346 472 472 191
Bidders 472 281 281 102
Panel B: Executive Characteristics
Rational 515 61.1% 178 63.3% 182 64.8% 68 66.7%
Overconfident 328 38.9% 103 36.7% 99 35.2% 34 33.3%
We can see that cash is used more frequently as a medium of payment in later deals (45.5% on 3rd deal) compared to first deals (36.7% on 1st deal) and that in all cases more than 60% of
14
acquisitions were within the same industry group. Moreover, overconfident CEO’s constitute at least one third of the overall CEO’s in the sample regardless of deal order.
Table 2 presents descriptive statistics of deals and acquiring firms. Panel A shows characteristics for the whole sample. Panel B, C and D show characteristics for the 1st, 2nd and 3rd acquisition respectively. Deal value is as reported to Thomson One. Relative Deal Value is the ratio of Deal Value to Assets. Market Cap is calculated as the product of shares outstanding and share price. Total debt is defined as long term debt plus debt in liabilities. Firm value is calculated as market value of equity plus total debt. Investment is equal to capital expenditures. Cash flow is calculated as sales less cost of goods sold. Tobin’s Q is defined as assets less book equity plus market equity, divided by book value of assets.
Table 2
This table shows descriptive statistics about the deals and acquiring firms. Panel A shows characteristics for the whole sample. Panel B, C and D show characteristics for subsamples of 1st, 2nd and 3rd acquisition respectively. Deal value is as reported to Thomson One. Relative Deal Value is the ratio of Deal Value to Assets. Assets is the book value of assets. Market Cap is calculated as the product of shares outstanding and share price. Total debt is defined as long term debt plus debt in liabilities. Firm value is calculated as market value of equity plus total debt. Investment is equal to capital expenditures. Cash flow is calculated as sales less cost of goods sold. Tobin’s Q is defined as assets less book equity plus market equity, divided by book value of assets. Values are in taken from Thomson One and the CRSP-Compustat database and are in millions of dollars, except ratios.
Acquirer Characteristics Panel A: All Deals
N Min Mean Median Max SD
Deal Value 1346 2 1403.1 275.66 22300.17 3358.61
Relative Deal Value 1346 0.002 0.37 0.08 5.65 0.85
Assets(Book) 1346 10.35 16982.98 3724.13 717000 46581.75
Market Cap 1343 19.73 17801.4 3872.88 180000 36013.06
Firm Value 1339 8.53 21070.52 5183.69 186000 39598.1
Cash & ST Investments 1346 0.53 1239.4 229.94 17760 2805.41
Cash/Assets 1346 0 0.12 0.06 0.68 0.14 Total Debt 1342 0 3380.4 863.7 43202 6665.17 Debt/Assets 1342 0 0.24 0.23 0.91 0.17 Investment 1339 0 626.56 132 10105 1436.5 Cash Flow 1343 11.85 3924.22 1051.24 43970 7991.42 Tobin's Q 1339 0.25 1.74 1.31 6.93 1.29
15
Panel B: 1st Deal
N Min Mean Median Max SD
Deal Value 472 2.38 940.76 187.78 18836.74 2327.26
Relative Deal Value 472 0.0021 0.31 0.07 5.65 0.72
Assets(Book) 472 10.35 14714.06 2850.56 717242.00 45626.07
Market Cap 471 19.73 12862.9 2835 180090.40 28556.84
Firm Value 470 8.53 15808.7 4173.78 185999.90 32486.74
Cash & ST Investments 472 0.53 946.9 144.62 17760.00 2459.76
Cash/Assets 472 0.0004 0.12 0.05 0.68 0.15 Total Debt 471 0 3026.41 639.73 43202.00 6593.41 Debt/Assets 471 0 0.25 0.24 0.91 0.17 Investment 469 0 535.14 110 10105.00 1356.66 Cash Flow 471 11.85 2963.37 843 43970.00 6669.20 Tobin's Q 470 0.25 1.66 1.25 6.93 1.22 Panel C: 2nd Deal
N Min Mean Median Max SD
Deal Value 472 2 1143.36 230.37 22300.17 2669.12
Relative Deal Value 472 0.0021 0.43 0.10 5.65 0.94
Assets(Book) 472 10.35 13930.65 2850.56 717242 43854.81
Market Cap 471 19.73 12422.14 2806.82 180090.40 27793.15
Firm Value 470 8.53 15226.18 4173.78 185999.90 31419.99
Cash & ST Investments 472 0.53 877.43 147.17 17760 2200.85
Cash/Assets 472 0.0004 0.12 0.05 0.68 0.15 Total Debt 471 0 2942.25 639.73 43202.00 6329.95 Debt/Assets 471 0 0.25 0.24 0.91 0.17 Investment 469 0 522.50 109.60 10105 1292.36 Cash Flow 471 11.85 2906.44 843.0 43970 6504.56 Tobin's Q 470 0.25 1.64 1.24 6.93 1.19 Panel D: 3rd Deal
N Min Mean Median Max SD
Deal Value 191 2 2065.64 417.55 22300.17 4550.49
Relative Deal Value 191 0.0021 0.38 0.09 5.65 0.88
Assets(Book) 191 94.67 22038.20 5629.96 717242 61025.61
Market Cap 190 91.43 20160.06 5001.49 180090.40 36875.14
Firm Value 189 150.22 24212.42 6962.03 185999.90 41280.65 Cash & ST Investments 191 0.87 1487.08 269.99 17760 3082.41
16 Total Debt 190 0 4214.98 1453.50 43202 7782.04 Debt/Assets 190 0 0.23 0.22 0.91 0.16 Investment 190 0 740.93 207.89 10105 1650 Cash Flow 190 32.17 4758.48 1401.55 43970 8893.47 Tobin's Q 189 0.25 1.70 1.31 6.93 1.26
We can see from Table 2 that the average deal value and firm size increases from the first deal toward the most recent ones. This makes sense intuitively, as with every acquisition the firm gets bigger and in turn it goes after bigger targets in the subsequent deals. For example, the median book value of assets increases from $2.85 billion in first deal to $5.629 billion on the last one, and the median deal value grows from $187 million to $417 million respectively. Furthermore, acquirers seem to be accumulating more debt as total debt grows from an average of $3 billion (1st deal) to $4.2 billion at the end of the acquisition series.
In Table 3, I present key descriptive statistics regarding the target firms of the sample. Panel A shows characteristics for the whole sample. Panel B, C and D show characteristics for subsamples of target for the 1st, 2nd and 3rd acquisition respectively. Assets is the book value of assets. Market Cap is calculated as the product of shares outstanding and share price. Total debt is defined as long term debt plus debt in liabilities. Firm value is calculated as market value of equity plus total debt. Investment is equal to capital expenditures. Cash flow is calculated as sales less cost of goods sold. In line with the results and interpretation of Table 2, we can see that indeed the targets of later acquisition rounds are larger than targets of first rounds. For instance, the median book value of assets has increased from $161.95 million to $233.43 million and the median firm value from $255 million to $309.25 million, from the first to third deal respectively.
17
Table 3
This table shows key characteristics of the target firms. Panel A shows characteristics for the whole sample. Panel B, C and D show characteristics for subsamples of firms for the 1st, 2nd and 3rd acquisition respectively. Assets is the book value of assets. Market Cap is calculated as the product of shares outstanding and share price. Total debt is defined as long term debt plus debt in liabilities. Firm value is calculated as market value of equity plus total debt. Investment is equal to capital expenditures. Cash flow is calculated as sales less cost of goods sold. Values are in taken from CRSP-Compustat database and are in millions of dollars, except ratios.
Target Characteristics Panel A: All Deals
N Min Mean Median Max SD
Assets (Book) 839 3.65 1345.42 187.33 115000 5353.96
Market Cap 833 5.07 1027.76 258.81 15567.19 2305.30
Firm Value 833 6.73 1318.79 318.15 18356.90 2856.66
Cash & ST Investments 839 0 81.65 17.70 1454.62 207.71
Cash/Assets 839 0 0.20 0.09 0.90 0.23 Total Debt 838 0 291.44 19.70 5248.38 796.02 Debt/Assets 838 0 0.20 0.14 0.85 0.20 Investment 820 0 56.69 7.55 857.10 138.30 Cash Flow 839 0.73 286.86 64.4 4728 686.75 Panel B: 1st Deal
N Min Mean Median Max SD
Assets (Book) 478 4.45 945.86 161.95 35637 2969.85
Market Cap 477 5.07 699.44 177.73 13133.44 1544.73
Firm Value 477 6.73 916.28 255.03 13786.65 1936.29
Cash & ST Investments 478 0 63.50 12.32 1454.62 177.13
Cash/Assets 478 0 0.17 0.07 0.90 0.21 Total Debt 478 0 209.44 22.56 5248.38 537.62 Debt/Assets 478 0 0.21 0.17 0.85 0.19 Investment 471 0 43.77 6.75 857.10 101.13 Cash Flow 478 0.73 223.68 53.47 4078 500.55 Panel C: 2nd Deal
N Min Mean Median Max SD
Assets (Book) 465 3.65 1004.57 185.74 30068 3123.76
Market Cap 464 5.07 868.25 269.70 15567.19 1868.78
Firm Value 464 6.73 1131.56 327.41 18356.90 2405.70
Cash & ST Investments 465 0 58.76 16.93 1454.62 147.16
Cash/Assets 465 0 0.18 0.07 0.90 0.22
18
Debt/Assets 465 0 0.21 0.16 0.85 0.21
Investment 460 0 48.78 7.78 857.10 122.00
Cash Flow 465 0.73 241.26 64.85 4728 579.89
Panel D: 3rd Deal
N Min Mean Median Max SD
Assets (Book) 192 6.68 2514.37 233.43 115447 11179.70
Market Cap 191 5.07 1431.33 283.51 15567.19 3151.10
Firm Value 191 6.73 1809.30 309.25 18356.90 3847.38
Cash & ST Investments 192 0 101.99 23.85 1454.62 256.76
Cash/Assets 192 0 0.20 0.09 0.88 0.23
Total Debt 191 0 399.68 28.77 5248.38 1016.52
Debt/Assets 191 0 0.19 0.12 0.82 0.19
Investment 199 0 70.32 9.23 857.10 165.24
Cash Flow 192 0.73 391.25 72.20 4728 937.79
An alternative interpretation of the data presented in Table 2 and Table 3, is consistent with the “eat or be eaten” theory of Gorton, Kahl & Rosen (2009) and neo-agency view of Phalippou et al. (2015). They argue that managers might acquire targets that do many acquisitions themselves, from fear that these targets can grow as much as to go after their own firms. So, they engage in acquisitions that can be described as “defensive”, driven by fear of losing their private benefits of control. That might be one of the reasons we notice targets of bigger size in later deals.
5. Results
5.1 CAR’s of serial acquirers
Using event study methodology I calculate the average abnormal returns across firms and by deal order and present them in Table 4. First, I find that across all deals made by frequent acquirers the mean cumulative abnormal returns are negative and statistically significant. Specifically, the three-day abnormal returns of bidders are on average -0.689% and -0.672% for first and second deal respectively, significant at 5% level, and -1, 24% for the third bid
19
announcement, significant at 1% level. The results are in line with (Fuller et al. 2002; Morck, Schleifer & Vishny, 1990; Croci & Petmezas, 2009, and Doukas & Petmezas 2007) who also report negative short-term abnormal returns, during acquisition announcement of public firms. These findings can be explained by the notion that overconfident or hubristic CEO’s, who decide on bad investment projects that result in shareholders wealth destruction, driven by completely different motives however.
Table 4
This table shows the evolution of cumulative average abnormal returns of acquiring firms stock (CAARi, t) from one deal to deal. Abnormal returns are calculated as actual returns less the predicted by the market model returns. The event window is from the day before to the day after the announcement of the bid. The estimation period is set to 200 trading days, ending 6 days prior to the announcement. I use t-test to conduct the statistical significance test. *, **, *** denotes significance at 10%, 5% and 1% level respectively.
Cumulative Abnormal Returns
1st Deal 2nd Deal 3rd Deal
Mean CAR's -0.689%** -0.672%** -1, 24%***
t-stat -1.98 -2.23 -2.73
Observations 472 472 191
*** p<0.01, ** p<0.05, * p<0.1
Second, it seems that abnormal returns for these acquirers decline in later deals, compared to the first ones which is along the same lines with the work of Ahern, 2008; Ismail, 2008; and Billet and Qian, 2008). Although from first to second deal there is a slight improvement from -0.689% to -0.672%, on the third deal returns decrease to a negative 1.24%. This decline is consistent with the notion of declining investment opportunities: the higher the deal order, the worse the returns will be, given that the most profitable projects are already implemented. Further, because the sample is made of public acquirers and targets, the more negative returns of last deals could be explained by the dilution of ownership in cases of stock payment. From the behavioral viewpoint, I do not find evidence of learning from past acquisitions. It can be argued that managers do not seem to learn from past performance to
20
implement the gained experience into their future decisions. The effect of overconfident CEO’s is tested further in section 5.2 using OLS and Logit regression models with an overconfidence proxy and several control variables.
5.2 Regression analysis
First, I run an OLS regression model to test the first hypothesis about the correlation of future abnormal returns and the effect of overconfidence. Results are presented in Table 5.
The cumulative abnormal returns of the most recent deal are regressed on the cumulative abnormal returns of the previous deal and the overconfidence proxy. The positive and significant coefficient of the “Previous CAR” suggests that indeed the cumulative abnormal returns are correlated from deal to deal. It remains significant after I control for several deal and acquirer/target characteristics, although it decreases from 0.115 (significant at 1% level) in models 1 & 2, to 0.0869 (significant at 10%) in model 3. I don’t not find evidence of overconfidence, as captured by the insignificant coefficient, affecting the abnormal returns upon acquisition announcement. Further, the inclusion of the interaction term is used to capture whether the CEO’s status as overconfident or rational has any effect on the correlation of past and future returns. Based on the insignificant coefficient of this term, I do not find evidence that overconfident managers causing announcement returns of past deal to have a stronger effect on the returns on future deals.
Table 5. OLS: Abnormal returns persistence.
This table shows the effect of previous deal announcement returns and overconfident CEO’s on the announcement returns of the current deal for acquirers. The three-day cumulative abnormal returns (CAR) are calculated as actual returns less predicted returns by the market model. The dependent variable is the CAR of the current (most recent) bid. Explanatory variables are: (i) the CAR from the previous deal; (ii) the overconfident proxy which equals one if the CEO holds exercisable stock options even though their value is at least 50% higher than value at grant date. Deal value is the logarithm of deal value as reported by Thomson One. Relative Deal Value defined as the ratio of Deal Value to acquirer’s book Assets. Only Cash is a dummy equal to one if the deal was financed with cash. Diversified is a dummy equal to one if the target and acquirer belong in different industry groups based on 2-digit Standard Industrial Classification (SIC) codes. Assets is the logarithm of book value of assets. Leverage is calculated as total debt over assets. Cash flow is the logarithm of sales less cost of goods sold. Tobin’s Q is defined as book assets less book equity plus market equity, divided by book value of assets. Market-to-Book is defined as firm value over book assets. *, **, *** denotes significance at 10%, 5% and 1% level respectively.
21 Current CAR (1) (2) (3) Previous CAR 0.115*** 0.115*** 0.0869* (3.329) (3.325) (1.660) Overconfident 0.00166 0.000545 (0.386) (0.108)
Previous CAR x Overconfident -0.0571
(-0.844)
Deal Value -0.00129
(-0.406)
Relative Deal Value -0.00175
(-0.452) Only Cash 0.00901* (1.722) Diversified -0.00258 (-0.480) Assets (Acquiror) 0.00320 (0.786) Cash/Assets (Acquiror) -0.0364 (-1.583) Leverage (Acquiror) 0.0334** (2.214)
Cash flow (Acquiror) -0.00369
(-0.840) Tobin’s Q (Acquiror) -0.0119*** (-5.676) Assets (Target) -0.00109 (-0.372) Cash/Assets (Target) -0.0110 (-0.833) Leverage (Target) -0.0501*** (-2.867) Market-to-Book (Target) 0.00210 (1.370) Constant -0.0131*** -0.0138*** 0.0208 (-6.029) (-4.549) (1.075) Observations 779 778 778 R-squared 0.016 0.016 0.151
Robust t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1
Next, I run a Logit regression model to test whether CEO's incorporate the negative signals sent by investors during their last acquisition, in their future decisions. Dependent variable is an indicator equal to one if current announcement returns are negative. Independent variable a dummy equal to one if previous announcement returns were negative “Previous CAR (Negative)”, and the overconfident proxy. Based on the learning hypothesis, managers are
22
supposed to learn from past performance and experience, and implement this new information when they decide on their next acquisition. For managers that had negative previous returns this would mean that by learning from the market’s reaction they would re-consider their future choice to achieve better results. I do not find, however, such evidence. The positive coefficient of “Previous CAR (Negative)” (statistically significant at 5% level in model 3) implies that firms with negative abnormal returns are more likely to continue having negative abnormal returns in the next bids. Moreover, the coefficients of overconfidence and the interaction terms are insignificant, suggesting that having an “irrational” CEO does not increase the likelihood of continuing having negative results in future bids, given negative results on previous deals. Overall the above is in contradiction with the learning theory. It seems that some CEO’s do not learn or adjust their decisions, being these CEO’s rational or overconfident. If learning holds, especially after a “bad” deal the management would change tactics as to perform better, and it would be more likely a “bad” deal to be followed by a “good” deal. These findings presented on Table 5 and Table 6, could be reconciled, by the investor sentiment effect as described by Zhu (2011) in the context of cross-border M&A. Investors may have optimistic views for future deals based on good past performance and pessimistic views in cases of bad performance. This could explain the correlation of acquisition announcement returns and the likelihood of bad deals followed by bad deals.
Table 6. Logit regressions
This table shows the likelihood of having negative abnormal returns in subsequent deals, given that abnormal returns were negative on the previous deal announcement and the CEO is overconfident. The three-day cumulative abnormal returns (CAR) are calculated as actual returns less predicted returns by the market model. The dependent variable is a dummy equal to one if current CAR is negative. Explanatory variables are: (i) A dummy equal to one if the previous deal CAR is negative; (ii) the overconfident proxy which equals one if the CEO holds exercisable stock options even though their value is at least 50% higher than value at grant date. Deal value is the logarithm of deal value as reported by Thomson One. Relative Deal Value defined as the ratio of Deal Value to acquirer’s book Assets. Only Cash is a dummy equal to one if the deal was financed with cash. Diversified is a dummy equal to one if the target and acquirer belong to a different industry group as based on 2-digit Standard Industrial Classification (SIC) codes. Assets is the logarithm of book value of assets. Leverage is calculated as total debt over assets. Cash flow is the logarithm of sales less cost of goods sold. Tobin’s Q is defined as book assets less book equity plus market equity, divided by book value of assets. Market-to-Book is defined as firm value over book assets. *, **, *** denotes significance at 10%, 5% and 1% level respectively.
23
Current CAR (Negative)
(1) (2) (3)
Previous CAR (Negative) 0.384*** 0.375*** 0.388**
(3.288) (3.198) (2.288)
Overconfident 0.0787 0.0625
(0.657) (0.387)
Previous CAR (Negative) x Overconfident 2.095
(1.045)
Deal Value 0.00464
(0.0506)
Relative Deal Value -0.127
(-1.056) Only Cash -0.183 (-1.117) Diversified -0.0170 (-0.104) Assets (Acquiror) 0.108 (0.905) Cash/Assets (Acquiror) 0.166 (0.240) Leverage (Acquiror) -0.482 (-0.892)
Cash flow (Acquiror) 0.183
(1.534) Tobin’s Q (Acquiror) 0.347*** (4.793) Assets (Target) 0.0124 (0.147) Cash/Assets (Target) 0.535 (1.196) Leverage (Target) 0.0168 (0.0347) Market-to-Book (Target) -0.00639 (-0.134) Constant -0.928*** -0.949*** -3.558*** (-10.80) (-9.608) (-6.302) Observations 1,353 1,346 819 Pseudo R2 0.00637 0.00623 0.0855 z-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1
Next, in Table 7 I use a Logit regression model to test this time if firms with overconfident managers are more likely to use cash as a method of payment. As Malmendier & Tate (2008) suggest, overconfident managers do not believe that the market values their firms fairly. As they overestimate their ability to create value they also have the perception that
24
the firm’s true valuation is higher than the market thinks. This leads them to view external capital as costly and are more willing to pay for acquisitions with internal cash rather than issue equity, especially if they run a cash rich firm. The model is estimated for the whole sample as well for the subsample of cash rich firms. As cash rich firms I define those that are on the top 25% of the sample based on their cash holdings. Across all specifications the dependent variable is a dummy equal to one if the deal was financed fully by cash. The explanatory variable is the overconfident proxy. I don’t not find any supporting evidence for the claim that overconfident managers are more likely to choose cash as medium of payment. The coefficient of interest in model (2), which tests the hypothesis for the whole sample, is not statistically different from zero. Moreover, in model (3), I run the same specification, but this time only for the subsample of firms with abundance in cash. The coefficient is again insignificant which suggests that the management’s overconfident behavior does not affect the payment method choice.
Table 7. Logit regressions
This table shows the effect of having an overconfident CEO on the likelihood of the acquisition being financed fully with cash. The dependent variable is a dummy equal to one if the deal was paid by cash. Explanatory variable of interest is the overconfident proxy which equals one if the CEO holds exercisable stock options even though their value is at least 50% higher than value at grant date. Deal value is the logarithm of deal value as reported by Thomson One. Relative Deal Value defined as the ratio of Deal Value to acquirer’s book Assets. Diversified is a dummy equal to one if the target and acquirer belong to a different industry group as based on 2-digit Standard Industrial Classification (SIC) codes. Assets is the logarithm of book value of assets. Leverage is calculated as total debt over assets. Cash flow is the logarithm of sales less cost of goods sold. Tobin’s Q is defined as book assets less book equity plus market equity, divided by book value of assets. Market-to-Book is defined as firm value over book assets. *, **, *** denotes significance at 10%, 5% and 1% level respectively.
Whole sample Whole sample Cash-rich
(1) (2) (3)
Overconfident -0.0142 -0.174 -0.490
(-0.125) (-1.107) (-1.544)
Deal Value -0.265*** -0.0985
(-2.873) (-0.505)
Relative Deal Value -0.131 -4.435**
(-0.937) (-1.967)
Diversified 0.139 -0.0425
(0.883) (-0.128)
Assets (Acquiror) -0.414*** -0.567*
25
Cash/Assets (Acquiror) -0.343 -0.655
(-0.501) (-0.405)
Leverage (Acquiror) -0.317 -1.536
(-0.598) (-1.085)
Cash flow (Acquiror) 0.644*** 0.695***
(4.950) (2.637) Tobin’s Q (Acquiror) -0.108 -0.174 (-1.534) (-1.304) Assets (Target) 0.0242 -0.0986 (0.287) (-0.633) Cash/Assets (Target) 0.225 -0.314 (0.506) (-0.376) Leverage (Target) -0.461 -0.947 (-0.976) (-0.935) Market-to-Book (Target) -0.0526 0.00150 (-1.073) (0.0204) Constant -0.285*** 0.483 2.326 (-4.072) (0.928) (1.130) Observations 1,346 828 232 Pseudo R2 8.52e-06 0.0794 0.176 z-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1 6. Conclusion
This study examines the potential effect of overconfident managers on the correlation between current acquisition announcement returns and the announcement returns of future bids, for frequent acquirers. It also examines the concept of learning through acquisition experience and the effect that overconfidence has on the method of payment. The empirical evidence derives from a sample of sequential deals of public U.S. bidder and targets between 1992 and 2016. First, the results are consistent with previous evidence of the literature, of negative abnormal returns for bidders that acquire public targets, and of the declining trend of these abnormal returns for serial acquirers. Second, I find a positive correlation between the returns of consecutive deals. However, I do not find any evidence that CEO’s overconfidence either strengthens or weakens this correlation. Overall the results can be interpreted as a contradiction to the learning hypothesis: neither rational, nor overconfident CEO’s seem to evaluate correctly the investors’ signals, reflected on the announcement of previous bids, to
26
“learn by experience” and achieve better results in their subsequent acquisitions. Finally, I find that the likelihood of choosing cash as the method of payment is not affected by overconfidence. Overconfident managers don’t seem to prefer to pay by cash, as suggested by the idea that external funds are perceived as costly due to under-valuation of their firms. I believe these findings may have important implications for corporate governance. If both rational and overconfident CEO’s don’t learn with experience, but continue to take value destructive actions, better control and monitoring mechanisms could help firms in achieving better results. Taking into account, however, that the sample is constructed with several selection criteria, the results could suffer from endogenous sample selection bias. Thus, I remain cautious about the generalization of these results to the entire firm-CEO population.
References
Ahern, K. R., Daminelli, D. & Fracassi, C., 2015. Lost in translation? The effect of cultural values on mergers around the world. Journal of Financial Economics, 117(1), pp. 165-189.
Aktas, N., Bodt, E. & Roll, R., 2010. Serial acquirer bidding: An empirical test of the learning hypothesis. Journal of Corporate Finance, Volume 17, pp. 18-32.
Andrade, G., Mitchell, M. & Stafford, E., 2001. New evidence and perspectives on mergers. Journal of Economic Perspectives, 15(2), pp. 103-120.
Betton, S., Eckbo, E. & Thoburn, K., 2008. Corporate takeovers. Handbook of Empirical Corporate Finance, Volume 2, pp. 291-485.
Billet , M. T. & Qian, Y., 2008. Are overconfident CEO's born or made? Evidence of self-attribution bias from frequent acquirers. Management Science, 54(6), pp. 1037-1051.
Boubakri, N., Chan, A. & Kooli, M., 2012. Are the busiest really the best? Further evidence from frequent acquirers. Journal of Multinational Financial Management, 22(1-2), pp. 1-23.
Cai, J. & Vijh, A. M., 2007. Inceptive effects of stock and option holdings of target and acquirer CEO's. The Journal of Finance, 62(4), pp. 1891-1933.
Cakici, N., Hessel, C. & Tandon, K., 1996. Foreign acquisitions in the United States: Effect on shareholder wealth of foreign acquiring firms. Journal of Banking & Finance, 20(2), pp. 307-329. Chang, S., 1998. Takeovers of privately held targets, methods of payment, and bidder returns. The Journal of Finance, 53(2), pp. 773-784.
27
Conn, R. L., Cosh, A., Guest, P. M. & Hughes, A., 2005. The impact on UK acquirers of domestic, cross-border, public and private acquisitions. Journal of Business Finance & Accounting, 32(5-6), pp. 815-870.
Croci, E. & Petmezas, D., 2009. Why do managers make serial acquisitions? An empirical investigation of performance predictability in serial acquisitions. Working Paper.
Doukas , J. A. & Petmezas, D., 2007. Acquisitions, overconfident managers and self-attribution bias. European Financial Management, 13(3), pp. 531-577.
Doukas, J., Holmen, M. & Travlos, N., 2002. Diversification, ownership and control of Swedish corporations. European Financial Management, 8(3), pp. 281-314.
Fee, E. & Thomas, S., 2004. Sources of gains in horizontal mergers: evidence from customer, supplier, and rival firms. Journal of Financial Economics, 74(3), pp. 423-460.
Fuller, K., Netter, J. & Stegemoller, M., 2002. What do returns to acquiring firms fell us? Evidence from firms that make many acquisitions. The Journal of Finance, 57(4), pp. 1763-1793.
Goergen, M. & Renneboorg, L., 2004. Shareholder wealth effect of European domestic and cross-border takeover bids. European Financial Management, 10(1), pp. 9-45.
Gorton, G., Kahl, M. & Rosen, R. J., 2009. Eat or be eaten: A theory of mergers and firm size. The Journal of Finance, 64(3), pp. 1291-1344.
Hackbarth, D. & Morellec, E., 2008. Stock returns in mergers and acquisitions. The Journal of Finance, 63(3), pp. 1213-1252.
Haleblian, J. J., Kim, J.-Y. & Rajagopalan, N., 2006. The influence of acquisition experience and performance on acquisition behavior: Evidence from the U.S. commercial naking industry. The Academy of Management Journal, 49(2), pp. 357-370.
Harford, J., 1999. Corporate cash reserves and acquisitions. The Journal of Finance, 54(6), pp. 1969-1997.
Harford, J., Jenter, D. & Li, K., 2011. Institutional cross-holdings and their effect on acquisition decisions. Journal of Financial Economics, 99(1), pp. 27-39.
Ismail, A., 2008. Which acquirers gain more, single or multiple?. Global Finance Journal, Volume 19, pp. 72-84.
Jensen, M. C., 1986. Agency costs of free cash flow, corporate finance and takeovers. The American Economic Review, 76(2), pp. 323-329.
Kohers, N. & Kohers, T., 2000. The value creation potential of high-tech mergers. Financial Analysts Journal, 56(3), pp. 40-50.
Malmendier, U. & Tate, G., 2008. Who makes acquisitions? CEO overconfidence and the market's reaction. Journal of Financial Economics, 89(1), pp. 20-43.
Maquieira, C. P., Megginson, W. L. & Nail, L., 1998. Wealth creation versus wealth redistributions in pure stock-for-stock mergers. Journal of Financial Economics, 48(1), pp. 3-33.
Martynova, M. & Renneboog, L., 2008. Spillover of corporate governance standards in cross-border mergers and acquisitions. Journal of Corporate Finance, 14(3), pp. 200-223.
28
Masulis, R. W., Wang, C. & Xie, F., 2007. Corporate governance and acquirer returns. The Journal of Finance, 62(4), pp. 1851-1889.
Moeller, S. B., Schlingemann, F. P. & Stulz, R. M., 2005. Firm size and the gains from acquisitions. Journal of Financial Economics, 73(2), pp. 201-228.
Morck, R., Shleifer, A. & Vishny, R. W., 1990. Do managerial objectives drive bad acquisitions?. The Journal of Finance, 45(1), pp. 31-48.
Myers, S. C. & Majluf, N. S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13(2), pp. 187-221. Officer, M. S., Poulsen, A. B. & Stegemoller, M., 2008. Target-firm information asymmetry and acquirer returns. Oxford University Press, 13(3), pp. 467-493.
Phalippou, L., Xu, F. & Zhao, H., 2015. Acquiring acquirers. Review of Finance, 19(4), pp. 1489-1541. Roll, R., 1986. The hubris hypothesis of corporate takeovers. The Journal of Business, 59(2), pp. 197-216.
Schipper, K. & Thompson, R., 1983. Evidence on the capitalized value of merger activity for acquiring firms. Journal of Financial Economics, 11(1-4), pp. 85-119.
Schwert, W., 1996. Markup pricing in mergers and acquisitions. Journal of Financial Economics, 41(2), pp. 153-192.
Shleifer, A. & Vishny, R. W., 1986. Large shareholders and corporate control. Journal of Political Economy, 94(3), pp. 461-488.
Travlos, N. G., 1987. Corporate takeover bids, methods of payments, and bidding firms' stock returns. The Journal of Finance, 42(4), pp. 943-963.
Zhu, P., 2011. Persistent performance and interaction effects in. Journal of Multinational Financial, Volume 21, pp. 18-39.
